The aim of this project is to develop a general seizure analysis and prediction system that employs a variety of advanced, Electroencephalogram (EEG) analysis and interpretation techniques implemented in a commercially viable and user-friendly research/clinical software package. Specifically, the software package will be developed (and tested clinically) for prediction and analysis of abnormal EEG patterns and seizures commonly seen in two distinct groups of patients: (1) Epileptic patients, and (2) Post cardiac-arrest patients with consequent interruption in brain perfusion resulting in cortical injury, seizure and abnormal EEG pattern. The specific aims of the project are: 1-To develop (and clinically test) a portfolio of novel linear/non-linear, time/frequency techniques for early and more precise detection of the time of seizure onset, 2-To utilize the above techniques for the development of techniques for a more precise determination of the region of seizure onset, 3- To implement the above techniques in a commercially viable and user-friendly research/clinical software package, and 4- to identify algorithms that are most suited for real-time seizure analysis/prediction implementation on a standard commercial PC. In a future follow-up work and under a separate grant, we plan to utilize these techniques as part of a portable seizure tracking and intervention system. [unreadable] [unreadable]